OpenRouter Configuration Guide

In this blog we are going to learn how to use OpenRouter for the AI module (Instant BI) of open source BI product Helical Insight.

Quick start – the whole process in 6 steps

In this guide

1. What is OpenRouter?

OpenRouter is a unified API gateway for large language models (LLMs). Instead of integrating each model provider separately, Helical Insight sends requests to OpenRouter using one API key and a model ID. This lets you try several supported models through a single credential and a common endpoint.

2. Before you begin

Make sure you have the following ready:

3. Create an OpenRouter API key

1. Go to openrouter.ai and sign in or create an account.
2. Open the Keys page from the dashboard, or go directly to openrouter.ai/keys.
3. Choose Create Key and give it a descriptive name, for example Helical-Insight-Dev.
4. If offered, set a spending limit or expiration that matches your organisation’s policy. YOu can also keep it 0 and use the free models.
5. Create the key and copy it immediately. OpenRouter shows the full secret only once, so save it in a password manager or secrets vault.
6. Review the pricing and usage limits for the model you plan to use.

4. Configure the OpenRouter provider

The provider file tells Instant BI how to talk to OpenRouter and which model to use initially.

1. In the same folder, open llm_config.yaml.
2. Add or update the openrouter block shown below.
3. Replace YOUR_OPENROUTER_API_KEY with your real key. Do this on the server only.
4. Save and close the file.

NOTE: Right now

llm_config.yaml

openrouter:
  package: langchain-openrouter
  model: nemotron-3.5-super-vl:free
  parameters:
    temperature: 0.1
    max_tokens: 4000
    api_key: YOUR_OPENROUTER_API_KEY
  usage_path: usage_metadata

What each setting does

Using several models?

Keep a valid default model in llm_config.yaml. The catalogue makes models available to choose from; the provider block defines the one configured first.

6. Select the provider in the UI

Sign in with an account that can manage system settings, then go to Management → Instant BI → LLM Providers. Menu wording can differ slightly between builds.

Figure 1. Instant BI → LLM Providers. The green tick marks the default provider.

1. Click the refresh icon if you just edited the YAML files.
2. Find the Openrouter tile. If it is missing, see Troubleshooting (Section 10).
3. Click the + / edit control on the tile to open its settings.
4. Pick the desired model from the Model dropdown.
5. Enter or verify the API key. Keep it private and keep it out of screenshots.
6. Review Temperature, Max Tokens, and Usage path.
7. Click Save.

Add models to the model catalogue in Helical Insight

Right now on the UI, even though “Open Router” icon is appearing, but it does not the models. Below steps explain how to configure the model names which will appear to be selected in the Open Router UI in Helical Insight.

The model catalogue file controls which models option appear in the UI.

1. On the Ubuntu server, open a terminal and go to the Instant BI configuration folder.
2. Open llm_models_catalog.yaml in a text editor.
3. Add langchain-openrouter entry and then add each model ID as a list item under it.
4. Save and close the file.

Terminal

cd ~/helical/helicalinsight-ee-7.0.1.278RC1/hi/hi-repository/System/InstantBI

nano llm_models_catalog.yaml

Example: one model

#llm_models_catalog.yaml

langchain-openrouter:
  - nemotron-3.5-super-vl:free

Example: multiple models

Put each model ID on its own line so every one appears in the UI.

#llm_models_catalog.yaml

langchain-openrouter:
  - nemotron-3.5-super-vl:free
  - another-model-id:free
  - third-model-id:free

Tips for a clean file

• Use spaces, not tabs, and keep the indentation consistent (two spaces before each -).
• Replace the example IDs with the exact IDs shown on OpenRouter. Do not assume a model stays available or free.

7. Set OpenRouter as the default provider

In the Edit Provider dialog, tick Set as default provider and click Save. A green tick on the provider tile confirms it is now the default.

Figure 2. Edit Provider (openrouter): model, default checkbox, API key, and parameters.

Just testing?

Leave Set as default provider unchecked to try OpenRouter without changing your current default. This works if your deployment supports it.

8. Switch between models

1. Confirm the model ID is listed under langchain-openrouter in llm_models_catalog.yaml.
2. Open the OpenRouter tile’s Edit Provider dialog.
3. Choose the required model from the Model dropdown.
4. Click Save, then confirm the selected model before you run a test.

Check model compatibility

Models differ. Some are text-only or may not support every capability an Instant BI workflow needs. Choose a model whose input/output types and context window fit your task.

Changes not showing?

Follow your deployment’s normal application reload or restart procedure, then check the Helical Insight logs for configuration or authentication errors.

9. Troubleshooting

10. Security best practices

• Treat OpenRouter API keys like passwords. Store them securely and restrict access to the config files.
• Never commit llm_config.yaml with a real key to source control or attach it to bug reports.
• If a key is exposed, revoke it and create a replacement.
• Recheck the live OpenRouter model page before deployment: free-model names, limits, and policies can change.
• Review the model and provider terms and data-handling policies before sending client or confidential data.

11. Quick reference and links

Key values

Official OpenRouter links